Problem with the k-omega model

Joachim · March 9, 2012, 05:32

Hello everyone!

I have a problem with a simulation, and I thought that maybe you could help me...
Actually, I am new at CFD, so this will maybe sounds simple to you. Anyway it isn't for me.

I am supposed to model a 2D airfoil (M, Re, AoA, etc given) with different turbulence models and then compare the results. First, I have established a grid convergence using Spalart Allmaras, and obtained very nice results in agreement with the experimental data. Then, I decided to try my fine mesh (150k cells) with the k-w model. The problem: the turbulent viscosity is too high, or the temperature...well, it doesn't work. Am I supposed to create a new mesh for the k-omega model?

I have uploaded some pictures of my mesh, do you see something wrong? There is y+=1 everywhere.

Thank you very much for your help!

Joachim

DarylMusashi · March 9, 2012, 14:56

Hello Joachim,

the expansion factor of the mesh shouldn't exceed 1.2, especially in the region near the blade - is that given? Furthermore why dont you use the SST-model instead of k-omega? The SST-model uses a switching function to give good results in the near-wall-regions (like k-omega does) and even far away from the walls (like k-epsilon does).

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem with turbulence model	akonduri	OpenFOAM	2	September 17, 2010 00:49
Turbulence model for mixing problem???	nileshjrane	Main CFD Forum	7	September 14, 2010 04:57
Problem in Herschel-Bulkely model	emad	FLUENT	0	May 20, 2009 14:20
Non premixed model - combustion validation problem	David	FLUENT	2	October 24, 2003 10:06
Two phase Eulerian Model Problem	senthil	FLUENT	5	May 21, 2003 02:55

March 9, 2012, 14:56		#2
DarylMusashi Senior Member Holger Dietrich Join Date: Apr 2011 Location: Germany Posts: 174 Rep Power: 14	Hello Joachim, the expansion factor of the mesh shouldn't exceed 1.2, especially in the region near the blade - is that given? Furthermore why dont you use the SST-model instead of k-omega? The SST-model uses a switching function to give good results in the near-wall-regions (like k-omega does) and even far away from the walls (like k-epsilon does).